Novel Application of Multiscale Cross-Approximate Entropy for Assessing Early Changes in the Complexity between Systolic Blood Pressure and ECG R-R Intervals in Diabetic Rats

Figures

• 

  Figure 1. Spontaneous oscillation of systolic blood pressure (SBP, gray line) and R-R intervals (RRI, black line). There is obscure similarity between the changes in SBP and RRI.

• 

  Figure 2. Consecutive systolic blood pressure (SBP1–4) and relevant R-R intervals (RRI1–3). According to the mechanism of baroreflex, the relevant RRI is defined as the consecutive RRI proceeded to the synchronously recorded SBP.

• 

  Figure 3. Multiscale cross-approximate entropy between systolic blood pressure and relevant R-R intervals from scales 1 to 10 in the normal controls (black circle), early diabetic rats in hyperglycemic state (white circle), and diabetic rats in euglycemic state (black triangle). *: significant difference between the controls and hyperglycemic rats; #: significant difference between the diabetic rats in hyperglycemic and euglycemic states. Significances of difference were determined as p < 0.05 by ANOVA with Fisher’s least significant difference post hoc test.

• 

  Figure 4. Spectral analyses of HRV among the normal controls (white) and diabetics in hyperglycemic (black) and treated euglycemic states (gray). *: Statistical significances were defined as p < 0.05 by ANOVA with Fisher’s least significant difference post hoc test.

• 

  Figure 5. Multiscale cross-approximate entropy of RRI-SBP raw data (black circle), surrogate shuffle series: shuffled paired RRI-SBP (shuffled paired, black triangle), and shuffled RRIs and shuffled SBP (separate shuffled, white circle). *: significant difference between the raw data and paired shuffle series; #: significant difference between the raw data and the separate shuffled series. Significances of difference were determined as p < 0.05 by ANOVA with Fisher’s least significant difference post hoc test.

• 

  Figure 6. The Bland–Altmann plots presents the 12 sets of agreements between the 3 types of entropy and 4 physiological parameters in the control rats. The 3 types of entropy are multiple scale cross-approximate entropy at small scale (scales 1–3; MS-CXApEnsmall), medium scale (scales 4–6; MS-CXApEnmedium), and large scale (scales 7–10; MS-CXApEnlarge). The 4 physiological parameters are heart rate variability by fast Fourier transform (LF/HF), Poincaré plotting (SSR), baroreflex sensitivity assessed by sequence method (BRSspn) and by pharmacological intervention (BRSphe). The agreements are acceptable between the 1.96 SD.